1. We have replicated high incidence of ventilator associated nosocomial pneumonia (VAP) in sheep, subjected to care similar to that now practiced in patients on mechanical ventilation in a hospital ICU: semirecumbent position with head elevated 30 , and periodic tracheal suctioning. Healthy sheep, anesthetized and paralyzed, were subjected to mechanical pulmonary ventilation in the prone position, with head, neck, and tracheal tube elevated 30 above horizontal, and suctioned periodically. Sheep developed progressive impairment of gas exchange, sepsis, and sometimes death. The trachea and the bronchopulmonary tree were uniformly heavily colonized with pathogens. Early studies, using continuous subglottic suctioning (Hi-Lo evac Mallinckrodt ETT), did not change this poor outcome. However, when sheep were positioned for 72 h prone, with trachea/tracheal tube horizontal and with no suctioning (routinely or otherwise), there was no growth of pathogens in the trachea and the give lobes of the lungs (sampled at 11 sites), normal lung function remained, and lungs appeared normal. We have also shown that the tracheal tube can be oriented horizontally with our custom gantry system, with periodic rotation of sheep from side to side, with equally good results. Importantly, nasogastric feeding was well tolerated (in contrast to conventional practice). We believe this gantry system and routine care based on our studies is readily adaptable to clinical use. 2. We have developed a thin-walled/two stage/lowest dead space/lowest resistance/twin endotracheal tube (UTTS-T-ETT) for use in adult, child, or infant. In the preterm infant, the dead space of this system is lower than that in the healthy, non-intubated newborn. Such a system can permit CPAP with least patient effort, and is likely to reduce greatly the need for mechanical ventilation in patients of all age groups.